Cryptococcus Neoformans Infection

New Insights on the Pathogenesis of Invasive Cryptococcus neoformans Infection

Helene C. Eisenman, PhD, Arturo Casadevall, MD, PhD, and Erin E. McClelland, PhD

Corresponding author Arturo Casadevall, MD, PhD lar mechanisms that kill yeast during transport through Albert Einstein College of Medicine, 411 Forchheimer, the blood, 3) infection of the central nervous system 1300 Morris Park Avenue, Bronx, NY 10461, USA. (CNS) by crossing the blood–brain barrier (BBB), and E-mail: [email protected] 4) growth in the brain [1,2•,3•]. We discuss how the Current Infectious Disease Reports 2007, 9:457–464 host and microbial factors are involved at each step in Current Medicine Group LLC ISSN 1523-3847 dissemination (Table 1). Copyright © 2007 by Current Medicine Group LLC

Growth in the Lung and Exit into the Disseminated cryptococcosis begins with infection of Bloodstream or Lymphatic System the lungs via inhalation. This is followed by escape from Host factors the lungs and entry into the bloodstream allowing dis- Containment of the microbe in the lungs is crucial in host semination to the brain and central nervous system. We defense against cryptococcosis. Much of the research on discuss the steps involved in dissemination and the host Cn pathogenesis in the lung relies on using the murine and microbial factors that influence each step. For the infection model. Mice are usually infected with Cn via host, containment in the lung is accomplished with a the intratracheal route to mimic the natural course of combination of cell-mediated and antibody responses. infection and generally succumb to cryptococcosis within Dissemination occurs when these systems fail and/or weeks of infection. In contrast, lung containment ulti- when phagocytic cells that fail to kill the yeast instead mately leads to reduced dissemination to the brain and act as a niche for replication. One of the main microbial other organs. Containment of Cn in the lung requires factors affecting dissemination is the polysaccharide an orchestrated immune response that relies on multiple capsule, a major virulence factor that promotes dissemi- components, including cell-mediated, antibody-based, nation at every step. Secreted enzymes are important, and innate immune mechanisms. including laccase and phospholipase B, which promote Cell-mediated immunity, including both CD8+ and escape from the lungs, and urease, which contributes CD4+ T cells, especially of the Th1 class, is required to to crossing the blood–brain barrier. Lastly, a number of eliminate Cn from the lungs. The production of Th1 cyto- regulatory factors contribute, especially to growth of kines and chemokines, especially for the recruitment and Cryptococcus neoformans in the brain. activation of macrophages, is essential. Macrophages play a key role in the immune defense against Cn. Containment of the fungal cells in granulomas in the lung is generally Introduction associated with a positive outcome for the host (Fig. 1). In For a host who has inhaled the opportunistic yeast contrast, a Th2 response is associated with dissemination Cryptococcus neoformans (Cn), the defining event is to the CNS and elsewhere with a negative outcome for the whether infection is cleared, remains confined to the host. Thus, there are many possible targets of immuno- lung, or undergoes extrapulmonary dissemination. The therapy in treating cryptococcosis. mechanism of dissemination is thought to involve the A number of Th1 cytokines are implicated in promot- following steps: 1) growth in the lung and crossing the ing a strong host defense in the lungs. For example, tumor bronchoalveolar epithelium to enter the lymphatic sys- necrosis factor (TNF)-α is produced in the lungs of infected tem for subsequent entry into the circulation or direct mice within a few days of infection. Inactivation with a invasion of the bloodstream, 2) escape from intravascu- single dose of anti–TNF-α antibody on the day of infec- 458 Fungal Infections

Table 1. Microbial and host factors involved in dissemination Stage of dissemination Microbial factors enhancing dissemination Host factors preventing dissemination Growth in lung/exit from lung into Capsule Th1 CD8+ T cells bloodstream or lymphatic system Laccase Th1 CD4+ T cells Phospholipase B TNF-a Inositol phosphosphingolipid-phospholipase IFN-g IL-18 Granulomas Antibodies Genetics of host Survival in the bloodstream or Capsule Macrophages lymphatic system Crossing the blood–brain barrier Transcytosis Macrophages Capsule Urease Growth in brain SRE1 Th1 cytokines CTR4 Th1 chemokines Genetics of host IFN—interferon; IL—interleukin; SRE—sterol regulatory element-binding protein; Th—T-helper; TNF—tumor necrosis factor.

Figure 1. Key factors affecting dissemination of Cn. A, Granuloma in the lung. Containment of the microbe in the lung is crucial to preventing disseminated infection. B, India ink stain of Cn showing the capsule surrounding the cell wall. The capsule has many immunomodulatory effects that protect the microbe from the host immune system. tion is sufficient to increase pulmonary fungal burden and in defense against Cn. When mice are infected with a highly dissemination of Cn to the brain and spleen [4]. Interferon virulent strain of Cn via the pulmonary route, administra- (IFN)-γ is another important Th1 cytokine. Inactivating tion of IL-18 improves survival and reduces fungal burden IFN-γ with antibodies worsens infection, as assessed by in the lung and brain. Also, it reduces the level of Cn cap- decreased survival and increased fungal burden in the lung sular glucuronoxylomannan (GXM) in the serum, another and brain. In contrast, administering recombinant IFN-γ indicator of systemic cryptococcosis [6]. improves the outcome of infection [5]. Similarly, interleu- What proves the importance of the Th1 response in kin (IL)-18, also called IFN-γ–inducing factor, is involved the lung is how it increases survival in those who receive Pathogenesis of Invasive Cryptococcus neoformans Infection Eisenman et al. 459 experimental treatments that activate it. A synthetic followed by later reactivation and dissemination. Evidence oligonucleotide containing an unmethylated CpG has suggests that exposure to Cn is common [14]. The organ- proven beneficial againstCn in mice. This oligonucleotide ism can remain latent in the body and later be reactivated is based on bacterial nucleotides shown to activate the by immunosuppressive therapies or disease. Rats provide immune response to several pathogens. CpG enhances the the best available model for the study of latency and reac- survival time and decreases the brain and lung fungal bur- tivation. Compared with mice, rats are less susceptible to dens of mice infected with Cn. The effect is via the Th1 cryptococcal disease. However, experimental infection response, as the Th1 cytokines IL-12 and IFN-γ increase with Cn shows that it can persist in the animals for more in bronchiolar lavage fluid, whereas IL-4 decreases. Also, than 1 year, surviving in host cells inside the lungs. Treat- the activity is dependent on CD4+ T cells [7]. Thus, cell- ment with the immunosuppressive agent dexamethasone mediated immunity, characterized by a Th1 response, is leads to increased growth and dissemination to other required for containing Cn in the lungs and ultimately organs [15]. Thus, the rodent model is useful for studying preventing dissemination. latent infection and reactivation, an important aspect of Antibodies are an important part of the immune the pathogenesis of Cn. response to Cn infection. Treating infected animals with The genetics and immune status of the host can greatly antibodies to GXM can enhance survival and reduce lung affect the dissemination of Cn. Certain host strains are fungal burdens and serum GXM [8]. In addition, human more susceptible to disseminated Cn infections. Testing of antibodies are effective against disseminated cryptococ- several common laboratory murine strains for susceptibil- cosis. Transgenic mice expressing human antibodies can ity to Cn shows that CBA/J mice are the most susceptible develop a protective antibody response to GXM mime- to infection and BALB/c the least, as determined by fun- topes and acquire immunity to Cn [9•]. However, the role gal burden in the brain. The difference is attributable to of antibody is complicated. B-cell–deficient mice have dissemination from the lungs, because no difference in decreased survival time and increased lung fungal bur- fungal burden is seen after intravenous infection, only dens compared with wild-type mice when infected with intratracheal infection. No single factor can explain the Cn. However, administration of antibodies against Cn difference in dissemination between the host strains. does not increase the survival of these mice, even though Instead, a combination of factors, including antibody it does in wild-type mice, indicating that passive antibody response, Th1/Th2 cytokine ratio, and the ability of Cn efficacy requires competent B cells. The B-cell–deficient to replicate inside host macrophages, likely cause the dif- mice have altered levels of inflammatory cytokines, sug- ference in dissemination of Cn in these hosts [16]. gesting alteration of the immune response. The decrease in survival time in these mice is therefore attributed to Microbial factors dysregulation of the immune response [10]. Numerous microbial virulence factors exist that promote The role of antibodies in lung pathogenesis of Cn can microbial growth in the lung and exit into the bloodstream be complicated by prozone-like effects. Some protective anti- or lymphatic system, including capsule, laccase, phospho- bodies against Cn GXM are disease enhancing at high doses, lipase B, and inositol phosphosphingolipid-phospholipase. and antibody efficacy varies with Cn inoculum. Antibodies The polysaccharide capsule (Fig. 1) is one of the major vir- that are protective with high inocula of Cn lose their protec- ulence factors of Cn and as such, has numerous inhibitory tive ability or become disease enhancing at a lower inoculum. effects on the host and the immune response. In addition, Thus, it seems that, at high doses, antibody may interfere the capsule is thought to give the yeast a growth advantage with certain immune functions, such as oxidative killing by in the lung because acapsular strains are generally cleared phagocytes. If this is the case, the high antibody dose might from the lung and rarely disseminate to the CNS [1]. increase phagocytosis but inhibit killing, thereby promot- The enzyme laccase catalyzes the first step of the path- ing the dissemination of Cn [11]. The effect of antibodies is way to produce melanin, another important virulence further complicated by the genetic background of the host. factor in Cn [2•]. When mice are infected intratrache- For example, IgG3 antibodies that are nonprotective in AJCr ally with either a laccase (-) or a laccase (+) strain, 70% and C57BL/6J mice are protective in C57BL/6J × 29Sv mice. of mice infected with the laccase (+) strain die by day This may be due to the antibodies having different effects in 30 postinfection compared with 100% survival in mice the context of different cell-mediated immune factors [12]. infected with the laccase (-) strain [2•]. Interestingly, Another example of antibody interaction with other immune there is no difference in pulmonary clearance, leukocyte components is its effect on the innate complement system. recruitment, inflammatory response, delayed-type hyper- Different monoclonal antibodies to GXM recognizing vari- sensitivity response, or cytokine production between the ous epitopes can alter the binding of C3 to the Cn capsule, laccase (-) and (+) strains. The only difference is seen in which may ultimately effect opsonization of Cn and/or the mice infected with the laccase (+) strain where there are inflammatory response [13]. higher numbers of Cn in extrapulmonary sites, which An aspect of cryptococcosis not studied in the systems suggests that laccase may be involved in dissemination. described above is long-term persistence of Cn in the lungs When the same strains are used to infect mice intrave- 460 Fungal Infections nously, there are equal numbers of both strains growing and ISC reconstituted strain, the isc mutant dissemi- in the brain and spleen after 1 week, suggesting that nates easily. The authors then tested the ability of the laccase is not involved in 1) survival and/or growth in strains to survive within activated macrophages, and extrapulmonary sites, 2) survival in the bloodstream not surprisingly, the isc mutant cannot survive after en route to extrapulmonary sites, or 3) survival within phagocytosis by an activated macrophage. Because the the brain. Instead, laccase seems to play a role in escape isc mutant is also extremely sensitive to acidic environ- from the lungs [2•]. ments and oxidative and nitrogen stress, the authors Phospholipase B is a secreted enzyme that removes acyl conclude that inositol phosphosphingolipid-phospholi- chains from phospholipids, such as those found in cellular pase C is important for allowing Cn to survive within membranes and lung surfactant, and is an important Cn macrophages in the lung [22]. virulence factor [17]. When mice are infected intratracheally with H99 (wild type), a plb mutant, and the PLB reconstituted strain, there is significantly higher fungal burden in Survival in Bloodstream or Lymphatic System extrapulmonary sites (including the brain) of H99 and the Host factors PLB reconstituted strain than the plb mutant strain, sug- Once Cn enters the circulatory system, it must survive for gesting a role for phospholipase in dissemination [18]. dissemination to occur. A potential mode of survival and Because the phospholipids in the lung surfactant are growth is inside macrophages. Macrophages have a dual preferred substrates of phospholipase B, the enzyme may role in the pathogenesis of Cn. They can phagocytose and initiate invasion of the lung interstitium and function in kill Cn, but this outcome does not always occur. Recent evi- exit from the lungs [19]. When mice are infected intratra- dence suggests that the susceptibility of a host may depend cheally with H99, a plb mutant, and the PLB reconstituted on the ability of macrophages to contain Cn [23]. Alter- strain, there is significantly less fungal burden associated nately, Cn can survive ingestion and intracellularly replicate. with lung interstitial macrophages from the plb mutant at Time-lapsed microscopy demonstrates that Cn replicates 24 hours, suggesting that phospholipase B facilitates the inside macrophages with a similar doubling time to fungal entry of Cn from the airways into the pulmonary inter- cells cultured in vitro. Intracellular growth of Cn is accom- stitium. Also, infection with the plb mutant is confined to panied by accumulation of polysaccharide vesicles into the the lung, compared with mice infected with H99 or the cytoplasm and eventual disruption of the phagolysosome. PLB reconstituted strain, which show infection in blood Fungal replication leads to eventual lysis and destruction of monocytes, plasma, peritoneal macrophages, cerebral the macrophages and release of live Cn cells [24]. In addi- tissue, and hilar lymph nodes. These data suggest that tion, phagocytosed Cn can exit infected cells by phagosomal phospholipase B is essential for hematogenous dissemina- extrusion without lysis, allowing both the host cell and tion of Cn from the lung and for lymphatic spread from pathogen to survive [25••]. This phenomenon appears to the lung to the regional lymph nodes [20•]. require specific host and microbial factors, including the Inositol phosphosphingolipid-phospholipase C is polysaccharide capsule [26]. Overall, these studies demon- involved in the degradation of inositol phosphoryl- strate that macrophages can be a vehicle for dissemination ceramide to phosphorylinositol and phytoceramide. for Cn by providing a site for reproduction and perhaps a Because inositol phosphorylceramide is involved in mechanism for distribution through the host. intracellular and extracellular fungal growth [21], the degradation of this sphingolipid is important in the Microbial factors regulation of fungal growth and pathogenicity. When GXM is the major component of the capsule and signifi- mice are infected intranasally with H99, an isc mutant, cant amounts of polysaccharide accumulate in serum. and the ISC reconstituted strain, there is significantly GXM is believed to contribute to cryptococcal virulence less fungal burden of the isc mutant compared with through its many deleterious effects on immune func- H99 and the ISC reconstituted strain in the brain, tion. This includes immune unresponsiveness, partly suggesting a role for inositol phosphosphingolipid- through the inhibition of T-cell activation and prolifera- phospholipase C in dissemination. The isc mutant is tion [27••,28], thus helping the organism escape killing only found extracellularly, is hyperencapsulated, and while in the bloodstream. consequently induces very little inflammation compared with H99 and the ISC reconstituted strain. Interest- ingly, when immunocompromised mice (Tgε26) lacking Crossing the BBB T and natural killer cells are infected with the three Host factors strains, the isc mutant has no problem disseminating For Cn to infect the CNS and cause meningoencephalitis, to the brain, which suggests that the host’s immune it must cross the BBB. One proposed mechanism is via response is important in preventing the isc mutant to infected phagocytic cells, also called the “Trojan Horse” disseminate. Indeed, when alveolar macrophages are mechanism. As discussed above, there is ample evidence depleted from mice infected with H99, the isc mutant, indicating that Cn can use macrophages as a vehicle for Pathogenesis of Invasive Cryptococcus neoformans Infection Eisenman et al. 461 replication. Thus, it is possible that Cn “hides” in mac- whereas only 30% of the mice infected with the urease (-) rophages to cross the BBB, and then escapes such cells by strain die over the entire 8-week experiment. Additionally, lysis or extrusion. the survivors have no visible evidence of Cn brain infection suggesting that deletion of urease prevents Cn infection/ Microbial factors pathology in the brain and prolongs host survival [3•]. One way the fungus can cross the BBB is by transcellular After intravenous, intratracheal, and intracerebral penetration through brain microvascular endothelial cells infection with H99, a urease (-) strain and the recon- (BMEC), or transcytosis. In vitro data provide compelling stituted urease (+) strain to determine where urease acts support for transcytosis. Cn cells can migrate across an in dissemination, there is no difference in organ fungal artificial BBB in a few hours. Furthermore, when Cn cells burden between the strains on days 1 and 4. This sug- are incubated with human BMECs in vitro, they penetrate gests that urease is not involved in survival in the brain, the cells after microvilli engulf them. Cn can also be seen lungs, spleen, or other extrapulmonary sites. When mice inside the BMECs, as well as exiting the opposite side of are infected intravenously, there is a 10-fold decrease the cells. In vivo, Cn cells associate with brain endothelium in fungal burden in organs with closed capillary beds within a few hours of infection [29••]. Thus, transcytosis (brain, lungs, heart, and kidney), compared with organs is a likely mechanism for Cn to cross the BBB, and there is with an open vasculature, such as the spleen, after 3 evidence that the polysaccharide capsule is involved [30]. hours. This finding is interpreted to suggest that urease Capsule changes resulting from phenotypic switch- enhanced yeast sequestration within the microcapillar- ing may help Cn cross the BBB. Phenotypic switching is ies of different organs. This finding may be related to defined as reversible colony morphology changes in theCn dissemination to the brain because histologic analysis capsule that arise in a population of cells at rates higher of the brain shows Cn within and in direct proximity than the background mutation rate [31]. Switch variants of the microcapillaries, suggesting that H99 invades show phenotypic differences in the morphology of the the brain via direct transfer across the BBB at the sites capsule, such as smooth, mucoid, and wrinkled, with the where Cn embolizes microcapillaries [3•]. mucoid phenotype generally having increased virulence over the smooth phenotype [32]. Recently, phenotypic switching has also been described for Cryptococcus gattii Growth in the Brain (serotype B) where switching is involved in dissemination. Host factors When mice are infected intratracheally with either smooth Similar host defense mechanisms are needed to contain Cn or mucoid variants, only smooth variants are isolated in the CNS as are needed in the lungs. When Cn is infected from the brain. Interestingly, smooth variants also have intracerebral into mice, Th1 cytokines and chemokines are smaller capsules than the mucoid variants. When mice produced in brains of infected animals [38]. Furthermore, are infected intravenously with either smooth or mucoid mice that produce a predominantly Th1 response (CBA/J) variants, mice infected with the smooth variant have sig- are more resistant than mice that produce a Th2 response nificantly higher fungal burden in the brain, suggesting (C57BL/6) after intracerebral infection [39]. that the smooth variant with the smaller capsule size is better at crossing the BBB [33]. Microbial factors Mutations affecting capsule size can influence the There are a few Cn factors that influence growth ability of Cn to cross the BBB. Capsule size is important in the brain. The sterol regulatory element-binding in crossing the BBB with an isc mutant where mutant cells protein (SRE1) plays a central role in adaptation to are hyperencapsulated and cannot infect the CNS [22]. low-oxygen growth in Schizosaccharomyces pombe Additionally, strains deficient in galactose metabolism [40] and has recently been implicated in virulence of such that the capsule contains no galactoxylomannan Cn. Interestingly, this mutant does not show a defect are also hyperencapsulated and cannot colonize the brain in the other steps of dissemination but is defective in after an intravenous infection in mice [34]. These data its ability to establish infection and grow in the brain. suggest that smaller capsulated cells may be more effec- Histologic data reveal that the sre mutant cells are tive in blood-brain penetration and is consistent with an limited to the outer regions of the brain (the leptomen- analysis showing a negative association between capsule inges and the cortex area) prompting the authors to size and virulence [35]. Thus, capsule size and changes conclude that sterol regulatory element-binding protein in polysaccharide structure appear to be important for is required for cell growth in the neuronal parenchyma crossing the BBB [36•]. and the white matter regions of the brain as well as The metalloenzyme urease catalyzes the hydrolysis for fulminating CNS infection. Because the sre mutant of urea to ammonia and carbamate and is an important is required for growth on nutrient-poor and low-iron virulence factor for Cn [37]. After intratracheal infection media, the limited growth in the brain may be due to of mice with the wild type H99 strain and a urease defi- differences in available nutrients among the leptomen- cient (-) strain, all mice infected with H99 die after week 6, inges, the cortex, and the neuronal parenchyma [41]. 462 Fungal Infections

Additionally, SRE1 controls the expression of CTR4p, Conclusions a high-affinity copper transporter that is also controlled To understand the pathogenesis of invasive infection by by the copper-dependent transcription factor CUF1. When Cn, it is imperative to discern how dissemination occurs. CUF1 is knocked out, the cuf1 mutant is avirulent in an If we can understand exactly what factors contribute to intravenous model of infection in mice and has signifi- each step of dissemination, it may be possible to develop cantly less fungal burden in the brain. Using a CTR4-GFP new therapies for patients with cryptococcosis. construct, the authors show that CTR4 is upregulated in Cn in low copper conditions. Using CTR4 expression as an assay, they find that intracellular Cn (within J774.16 Acknowledgment cells) and fungal cells isolated from the mouse brain have Dr. Eisenman and Dr. McClelland contributed equally to increased levels of CTR4, suggesting low copper availabil- this work. ity. Interestingly, CTR4 expression from isolates of solid organ transplant patients with Cn infection is increased in patients with CNS infection compared with patients with References and Recommended Reading a pulmonary infection. Thus, CTR4 expression (and low Papers of particular interest, published recently, copper availability) is highly correlated with Cn growth have been highlighted as: in the brain [42•]. Interestingly, in the sre1 mutant strain, • Of importance CTR4 expression is upregulated, providing an additional •• Of major importance explanation of low copper availability for poor growth in 1. Wilder JA, Olson GK, Chang YC, et al.: Complementation the brain [41]. of a capsule deficient Cryptococcus neoformans with CAP64 restores virulence in a murine lung infection. Am J Respir Cell Mol Biol 2002, 26:306–314. 2.• Noverr MC, Williamson PR, Fajardo RS, Huffnagle GB: Cryptococcosis and AIDS CNLAC1 is required for extrapulmonary dissemination of Cn is an opportunistic pathogen. Normally, the host Cryptococcus neoformans but not pulmonary persistence. Infect Immun 2004, 72:1693–1699. immune response is adequate to contain or eliminate The first publication to dissect at which step of dissemination the microbe. However, under certain conditions, Cn can laccase is involved. escape and disseminate throughout the host. This involves 3.• Olszewski MA, Noverr MC, Chen GH, et al.: Urease a number of microbial and host factors that seem to be expression by Cryptococcus neoformans promotes microvascular sequestration, thereby enhancing central nervous especially important in AIDS, as many aspects of HIV system invasion. Am J Pathol 2004, 164:1761–1771. infection create a situation in which the host is suscepti- The first publication showing how Cn uses urease to cross the BBB. ble. Thus, disseminated cryptococcosis is a very common 4. Huffnagle GB, Toews GB, Burdick MD, et al.: Affer- ent phase production of TNF-alpha is required for the disease in patients with AIDS, especially in the developing development of protective T cell immunity to Cryptococcus world where meningitis caused by Cn infection is found in neoformans. J Immunol 1996, 157:4529–4536. up to 45% of patients with AIDS [43], mainly due to lack 5. Kawakami K, Tohyama M, Teruya K, et al.: Contribution of interferon-gamma in protecting mice during pulmonary of access to highly active antiretroviral therapy (HAART) and disseminated infection with Cryptococcus neoformans. [44]. This is unfortunate, as histology studies have shown FEMS Immunol Med Microbiol 1996, 13:123–130. that AIDS patients given HAART have a better immune 6. Kawakami K, Qureshi MH, Zhang T, et al.: IL-18 protects response against Cn infection and that patients without mice against pulmonary and disseminated infection with Cryptococcus neoformans by inducing IFN-gamma produc- access to HAART offer little resistance to dissemination tion. J Immunol 1997, 159:5528–5534. by Cn [45]. 7. Miyagi K, Kawakami K, Kinjo Y, et al.: CpG oligodeoxy- From the perspective of the host, HIV infection is nucleotides promote the host protective response against + infection with Cryptococcus neoformans through induction associated with a reduction in CD4 T cells, producing of interferon-gamma production by CD4+ T cells. Clin Exp profound defects in cellular immunity that presumably Immunol 2005, 140:220–229. markedly increase susceptibility and dissemination 8. Feldmesser M, Casadevall A: Effect of serum IgG1 to Cryp- tococcus neoformans glucuronoxylomannan on murine [46]. One defect is a shift toward a Th2 response that pulmonary infection. J Immunol 1997, 158:790–799. inhibits cellular immune responses that block dissemi- 9.• Maitta RW, Datta K, Chang Q, et al.: Protective and nation [47], as seen in AIDS patients with disseminated nonprotective human immunoglobulin M monoclonal anti- Cn infection who have significantly higher levels of the bodies to Cryptococcus neoformans glucuronoxylomannan manifest different specificities and gene use profiles. Infect Th2 cytokines TNF-α and IL-10 [48]. Additionally, HIV Immun 2004, 72:4810–4818. infection of alveolar macrophages impairs the patients’ This paper shows that human antibodies to GXM are protective innate fungicidal activity against Cn [49]. Not only does in mice. 10. Rivera J, Zaragoza O, Casadevall A: Antibody-mediated AIDS increase susceptibility to cryptococcosis, but Cn protection against Cryptococcus neoformans pulmonary may also influence the course of HIV infection because infection is dependent on B cells. Infect Immun 2005, GXM has been shown to enhance the infectivity of HIV 73:1141–1150. 11. Taborda CP, Rivera J, Zaragoza O, Casadevall A: More in susceptible cells, probably due to enhancement of the is not necessarily better: prozone-like effects in passive viral membrane protein gp120 binding to CD4 [50]. immunization with IgG. J Immunol 2003, 170:3621–3630. Pathogenesis of Invasive Cryptococcus neoformans Infection Eisenman et al. 463

12. Rivera J, Casadevall A: Mouse genetic background is 29.•• Chang YC, Stins MF, McCaffery MJ, et al.: Cryptococcal a major determinant of isotype-related differences for yeast cells invade the central nervous system via transcel- antibody-mediated protective efficacy against Cryptococcus lular penetration of the blood-brain barrier. Infect Immun neoformans. J Immunol 2005, 174:8017–8026. 2004, 72:4985–4995. 13. Kozel TR, deJong BC, Grinsell MM, et al.: Characteriza- One of the first publications showing how Cn crosses the BBB. tion of anticapsular monoclonal antibodies that regulate 30. Chen SH, Stins MF, Huang SH, et al.: Cryptococcus neo- activation of the complement system by the Cryptococcus formans induces alterations in the cytoskeleton of human neoformans capsule. Infect Immun 1998, 66:1538–1546. brain microvascular endothelial cells. J Med Microbiol 14. Goldman DL, Khine H, Abadi J, et al.: Serologic evidence 2003, 52:961–970. for Cryptococcus neoformans infection in early childhood. 31. Guerrero A, Jain N, Goldman DL, Fries BC: Phenotypic Pediatrics 2001, 107:E66. switching in Cryptococcus neoformans. Microbiology 15. Goldman DL, Lee SC, Mednick AJ, et al.: Persistent 2006, 152:3–9. Cryptococcus neoformans pulmonary infection in the rat is 32. Fries BC, Taborda CP, Serfass E, Casadevall A: Phenotypic associated with intracellular parasitism, decreased induc- switching of Cryptococcus neoformans occurs in vivo and ible nitric oxide synthase expression, and altered antibody influences the outcome of infection. J Clin Invest 2001, responsiveness to cryptococcal polysaccharide. Infect Immun 108:1639–1648. 2000, 68:832–838. 33. Jain N, Li L, McFadden DC, et al.: Phenotypic switching in 16. Zaragoza O, Alvarez M, Telzak A, et al.: The Relative a Cryptococcus neoformans variety gattii strain is associated Susceptibility of Mouse Strains to Pulmonary Cryptococ- with changes in virulence and promotes dissemination to the cus neoformans Infection Is Associated with Pleiotropic central nervous system. Infect Immun 2006, 74:896–903. Differences in the Immune Response. Infect Immun 2007, 34. Moyrand F, Fontaine T, Janbon G: Systematic capsule gene 75:2729–2739. disruption reveals the central role of galactose metabolism 17. Chen SC, Muller M, Zhou JZ, et al.: Phospholipase activity on Cryptococcus neoformans virulence. Mol Microbiol in Cryptococcus neoformans: a new virulence factor? J Infect 2007, 64:771–781. Dis 1997, 175:414–420. 35. McClelland EE, Perrine WT, Potts WK, Casadevall A: 18. Noverr MC, Cox GM, Perfect JR, Huffnagle GB: Role The relationship of virulence factor expression to evolved of PLB1 in pulmonary inflammation and cryptococcal virulence in mouse-passaged Cryptococcus neoformans eicosanoid production. Infect Immun 2003, 71:1538–1547. lines. Infect Immun 2005, 73:7047–7050. 19. Santangelo RT, Nouri-Sorkhabi MH, Sorrell TC, et al.: 36.• Charlier C, Chretien F, Baudrimont M, et al.: Capsule Biochemical and functional characterisation of secreted phos- structure changes associated with Cryptococcus neofor- pholipase activities from Cryptococcus neoformans in their mans crossing of the blood-brain barrier. Am J Pathol naturally occurring state. J Med Microbiol 1999, 48:731–740. 2005, 166:421–432. 20.• Santangelo R, Zoellner H, Sorrell T, et al.: Role of extracel- A paper that definitively shows that the capsule changes when lular phospholipases and mononuclear phagocytes in crossing the BBB. dissemination of cryptococcosis in a murine model. Infect 37. Cox GM, Mukherjee J, Cole GT, et al.: Urease as a viru- Immun 2004, 72:2229–2239. lence factor in experimental cryptococcosis. Infect Immun The first publication to describe how phospholipase is used by Cn 2000, 68:443–448. in dissemination. 38. Uicker WC, Doyle HA, McCracken JP, et al.: Cytokine 21. Luberto C, Toffaletti DL, Wills EA, et al.: Roles for inosi- and chemokine expression in the central nervous system tol-phosphoryl ceramide synthase 1 (IPC1) in pathogenesis associated with protective cell-mediated immunity against of C. neoformans. Genes Dev 2001, 15:201–212. Cryptococcus neoformans. Med Mycol 2005, 43:27–38. 22. Shea JM, Kechichian TB, Luberto C, Del Poeta M: The cryp- 39. Huffnagle GB, McNeil LK: Dissemination of C. neoformans tococcal enzyme inositol phosphosphingolipid-phospholipase to the central nervous system: role of chemokines, Th1 C confers resistance to the antifungal effects of macrophages immunity and leukocyte recruitment. J Neurovirol 1999, and promotes fungal dissemination to the central nervous 5:76–81. system. Infect Immun 2006, 74:5977–5988. 40. Hughes AL, Todd BL, Espenshade PJ: SREBP pathway 23. Shao X, Mednick A, Alvarez M, et al.: An innate immune responds to sterols and functions as an oxygen sensor in system cell is a major determinant of species-related fission yeast. Cell 2005, 120:831–842. susceptibility differences to fungal pneumonia. J Immunol 41. Chang YC, Bien CM, Lee H, et al.: Sre1p, a regulator of 2005, 175:3244–3251. oxygen sensing and sterol homeostasis, is required for 24. Tucker SC, Casadevall A: Replication of Cryptococcus virulence in Cryptococcus neoformans. Mol Microbiol neoformans in macrophages is accompanied by phagosomal 2007, 64:614–629. permeabilization and accumulation of vesicles containing 42.• Waterman SR, Hacham M, Hu G, et al.: Role of a CUF1/ polysaccharide in the cytoplasm. Proc Natl Acad Sci U S A CTR4 copper regulatory axis in the virulence of Cryptococ- 2002, 99:3165–3170. cus neoformans. J Clin Invest 2007, 117:794–802. 25.•• Ma H, Croudace JE, Lammas DA, May RC: Expulsion One of the first publications showing that nutrient accessibility is of live pathogenic yeast by macrophages. Curr Biol 2006, important for Cn to grow in the brain. 16:2156–2160. 43. Hakim JG, Gangaidzo IT, Heyderman RS, et al.: Impact In concordance with Alvarez et al. [26], this paper demonstrates of HIV infection on meningitis in Harare, Zimbabwe: a that Cn can replicate inside macrophages and subsequently be prospective study of 406 predominantly adult patients. expelled without macrophage lysis. AIDS 2000, 14:1401–1407. 26. Alvarez M, Casadevall A: Phagosome Extrusion and Host- 44. Heyderman RS, Gangaidzo IT, Hakim JG, et al.: Cell Survival after Cryptococcus neoformans Phagocytosis Cryptococcal meningitis in human immunodeficiency by Macrophages. Curr Biol 2006, 16:2161–2165. virus-infected patients in Harare, Zimbabwe. Clin Infect 27.•• Yauch LE, Lam JS, Levitz SM: Direct Inhibition of T-Cell Dis 1998, 26:284–289. responses by the cryptococcus capsular polysaccharide 45. Shibuya K, Hirata A, Omuta J, et al.: Granuloma and glucuronoxylomannan. PLoS Pathog 2006, 2:e120. cryptococcosis. J Infect Chemother 2005, 11:115–122. This paper demonstrates that T-cell proliferation in response 46. Huffnagle GB, Yates JL, Lipscomb MF: T cell-mediated to various antigens is inhibited directly by GXM, a novel immunity in the lung: a Cryptococcus neoformans pulmo- virulence mechanism. nary infection model using SCID and athymic nude mice. 28. Vecchiarelli A: Immunoregulation by capsular components Infect Immun 1991, 59:1423–1433. of Cryptococcus neoformans. Med Mycol 2000, 38:407–417. 464 Fungal Infections

47. Cherniak R, Sundstrom JB: Polysaccharide antigens of the 49. Ieong MH, Reardon CC, Levitz SM, Kornfeld H: Human capsule of Cryptococcus neoformans. Infect Immun 1994, immunodeficiency virus type 1 infection of alveolar macro- 62:1507–1512. phages impairs their innate fungicidal activity. Am J Respir 48. Lortholary O, Improvisi L, Rayhane N, et al.: Cytokine Crit Care Med 2000, 162:966–970. profiles of AIDS patients are similar to those of mice with 50. Pettoello-Mantovani M, Casadevall A, Kollmann TR, et al.: disseminated Cryptococcus neoformans infection. Infect Enhancement of HIV-1 infection by the capsular polysaccha- Immun 1999, 67:6314–6320. ride of Cryptococcus neoformans. Lancet 1992, 339:21–23.